Update README.md

README.md

@@ -51,7 +51,7 @@ dataset_summary: '
 
   # Note: other available arguments include ''max_samples'', etc
 
-  dataset = fouh.load_from_hub("rohis/DUTS")
+  dataset = fouh.load_from_hub("Voxel51/DUTS")
 
 
   # Launch the App
@@ -102,11 +102,10 @@ session = fo.launch_app(dataset)
 
 ### Dataset Description
 
-<!-- Provide a longer summary of what this dataset is. -->
+DUTS is a saliency detection dataset containing 10,553 training images and 5,019 test images. All training images are collected from the ImageNet DET training/val sets, while test images are collected from the ImageNet DET test set and the SUN data set. Both the training and test set contain very challenging scenarios for saliency detection. Accurate pixel-level ground truths are manually annotated by 50 subjects.
 
 
-
-- **Curated by:** Introduced by Wang et al. in [Learning to Detect Salient Objects With Image-Level Supervision](https://paperswithcode.com/paper/learning-to-detect-salient-objects-with-image)
+- **Curated by:** Lijun Wang, Huchuan Lu, Yifan Wang, Mengyang Feng, Dong Wang, Baocai Yin, and Xiang Ruan 
 - **Language(s) (NLP):** en
 - **License:** unknown